info.debatty.java.stringsimilarity

Class Cosine

java.lang.Object

info.debatty.java.stringsimilarity.Cosine

All Implemented Interfaces:

StringSimilarityInterface

public class Cosine
extends Object
implements StringSimilarityInterface

Implements Cosine Similarity. The strings are first transformed in vectors of occurences of k-shingles (sequences of k characters). In this n-dimensional space, the similarity between the two strings is the cosine of their respective vectors.

Author:

Thibault Debatty

Constructor Summary

Constructors

Constructor and Description
Cosine()
Cosine(int k)

Method Summary

Methods

Modifier and Type	Method and Description
double	distance(String s1, String s2) Generally, distance = 1 - similarity.
protected static double	dotProduct(int[] v1, int[] v2)
static void	main(String[] args)
protected static double	norm(int[] v) Compute the norm L2 : sqrt(Sum_i( v_i^2))
double	similarity(String s1, String s2) Computes the cosine similarity of s1 and s2.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

Cosine

public Cosine(int k)

Cosine

public Cosine()

Method Detail

main

public static void main(String[] args)

Parameters:

args - the command line arguments

similarity

public double similarity(String s1,
                String s2)

Computes the cosine similarity of s1 and s2. The strings are first converted to vectors in the space of k-shingles. The cosine similarity is computed as V1 . V2 / (|V1| * |V2|)

Specified by:

similarity in interface StringSimilarityInterface

Parameters:

s1 -

s2 -

Returns:

Cosine similarity

distance

public double distance(String s1,
              String s2)

Description copied from interface: StringSimilarityInterface

Generally, distance = 1 - similarity. Some implementations can also provide a method distanceAbsolute

Specified by:

distance in interface StringSimilarityInterface

Returns:

distance between 0 (s1 = s2) and 1 (completely different)

norm

protected static double norm(int[] v)

Compute the norm L2 : sqrt(Sum_i( v_i^2))

Parameters:

v -

Returns:

L2 norm

dotProduct

protected static double dotProduct(int[] v1,
                int[] v2)